Data Processing Engine Research Articles

A learned cost model for big data query processing

The efficiency of query processing in the Spark SQL big data processing engine is significantly affected by execution plans and allocated resources. However, existing cost models for Spark SQL rely on hand-crafted rules. While learning-based cost models have been proposed for relational databases, they do not consider available resources. To address this issue, we propose a resource-aware deep learning model capable of automatically predicting query plan execution times based on historical data. To train our model, we embed query execution plans within a query plan tree and extracted features from allocated resources. An adaptive attention mechanism is integrated into the deep learning model to enhance prediction accuracy. Additionally, we extract sufficient features to represent data information and learn the effect of the data on query execution. This approach reduces the need for model retraining owing to data changes. The experimental results demonstrate that our deep cost model outperforms traditional rule-based methods and relational database learning-based optimizers in predicting query plan execution times.

Data Processing Engine (DPE): data analysis tool for particle tracking and mixed radiation field characterization with pixel detectors Timepix

Hybrid semiconductor pixelated detectors from the Timepix family are advanced detectors for online particle tracking, offering energy measurement and precise time stamping capabilities for particles of various types and energies. This inherent capability makes them highly suitable for various applications, including imaging, medical fields such as radiotherapy and particle therapy, space-based applications aboard satellites and the International Space Station, and industrial applications. The data generated by these detectors is complex, necessitating the development and deployment of various analytical techniques to extract essential information.For this purpose, and to aid the Timepix user community, it was designed and developed the“Data Processing Engine” (DPE) as an advanced tool for data processing designed explicitly for Timepix detectors. The functionality of the DPE is structured into three distinct processing levels: i) Pre-processing: this phase involves clusterization and the application of necessary calibrations and corrections. ii) Processing: this stage includes particle classification, employing machine learning algorithms, and the recognition of radiation fields. iii) Post-processing: involves various analyses, such as directional analysis, coincidence analysis, frame analysis, Compton directional analysis, and the generation of physics products, are performed. The core of the DPE is supported by an extensive experimental database containing calibrations and referential radiation fields of typical environments, including protons, ions, electrons, gamma rays and X rays, as well as thermal and fast neutrons. To enhance accessibility, the DPE is implemented into various user interface platforms such as a command-line tool, an application programming interface, and as a graphical user interface in the form of a web portal. The DPE's broad utility is exemplified through its integration into various applications and developments.

Leveraging Data Processing Engine for ISV: Building, Packaging, and Distributing Industry - Specific Solutions

Leveraging Data Processing Engine for ISV: Building, Packaging, and Distributing Industry - Specific Solutions

Machine learning based data collection protocol for intelligent transport systems: a real-time implementation on Dublin M50, Ireland

The continuous global urbanization with rapid and dynamic transitioning in traffic situations among highly populated cities results in difficulty for data collection and communication. Data collection for millions of vehicles hinders by various problems, i.e., higher cost of energy, time, space, and storage resources. Moreover, higher data traffic results in higher delays, larger throughput, excessive bottlenecks, and frequent repetition of data. To better facilitate the aforementioned challenges and to provide a solution, we have proposed a lightweight Machine Learning based data collection protocol named ML-TDG to effectively deal with higher data volumes in a real-time traffic environment capable of bringing the least burden on the network while utilizing less space, time, and energy. ML-TDG is functional based on Apache Spark, an effective data processing engine that indexes the data based on two logs, i.e., old commuters or frequent/daily commuters and second new/occasional commuters. The proposed protocol’s main idea is to utilize real-time traffic, distinguish the indexes in parallel based on two assigned logs criteria to train the network, and collect data with the least sources. For energy and time optimization, dynamic segmentation switching is introduced which is an intelligent road segments division and switching for reducing bottlenecks and replication. ML-TDG is tested and verified on Dublin, Ireland’s busiest motorway M50. ML-TDG performs the data collection, data sorting, and network training to decide the next execution altogether for better optimization every time. The experimental results verify that our proposed protocol is attaining higher performance with lower resource requirements along with rich and time-efficient sustainable data collection clusters in comparison with baseline protocols.

Efficient Large-Scale GPS Trajectory Compression on Spark: A Pipeline-Based Approach

Every day, hundreds of thousands of vehicles, including buses, taxis, and ride-hailing cars, continuously generate GPS positioning records. Simultaneously, the traffic big data platform of urban transportation systems has already collected a large amount of GPS trajectory datasets. These incremental and historical GPS datasets require more and more storage space, placing unprecedented cost pressure on the big data platform. Therefore, it is imperative to efficiently compress these large-scale GPS trajectory datasets, saving storage cost and subsequent computing cost. However, a set of classical trajectory compression algorithms can only be executed in a single-threaded manner and are limited to running in a single-node environment. Therefore, these trajectory compression algorithms are insufficient to compress this incremental data, which often amounts to hundreds of gigabytes, within an acceptable time frame. This paper utilizes Spark, a popular big data processing engine, to parallelize a set of classical trajectory compression algorithms. These algorithms consist of the DP (Douglas–Peucker), the TD-TR (Top-Down Time-Ratio), the SW (Sliding Window), SQUISH (Spatial Quality Simplification Heuristic), and the V-DP (Velocity-Aware Douglas–Peucker). We systematically evaluate these parallelized algorithms on a very large GPS trajectory dataset, which contains 117.5 GB of data produced by 20,000 taxis. The experimental results show that: (1) It takes only 438 s to compress this dataset in a Spark cluster with 14 nodes; (2) These parallelized algorithms can save an average of 26% on storage cost, and up to 40%. In addition, we design and implement a pipeline-based solution that automatically performs preprocessing and compression for continuous GPS trajectories on the Spark platform.

Apache Flink and clustering-based framework for fast anonymization of IoT stream data

In this paper, we present a novel framework that considers the expiration period time of the Internet of Things (IoT) data stream to anonymize it. IoT stands among one of most fast-growing technology in the world. Also, anonymity is one of the safeguards in place to protect data privacy. Because of the dynamic nature, vastness, and rapid changes in data streams, traditional approaches cannot be used to anonymize IoT data. The anonymization framework proposed in this paper performs its operation using a new clustering method and Apache Flink flow data processing engine. In this framework, firstly, we cluster received data. Then, if the size of the clusters doesn't meet the K-anonymity threshold, our review will continue to suppress and delete them; otherwise, the data would be anonymized and published. In this way, the framework handles both numerical and categorical data. At the end of the stream, the final remaining data will be merged and anonymized. Implementing and evaluating the framework using Scala and Apache Flink shows that the proposed approach reduces data delay by 12.33–66.62% compared with the other methods. Furthermore, in the end, combining the leftover clusters avoids information loss. In comparison with similar methods, information loss is reduced by 5.68–18.26%. The evaluation results show better performance in terms of data delay and information loss.

Big Data Analytic Toolkit: A General-Purpose, Modular, and Heterogeneous Acceleration Toolkit for Data Analytical Engines

Query compilation and hardware acceleration are important technologies for optimizing the performance of data processing engines. There have been many works on the exploration and adoption of these techniques in recent years. However, a number of engines still refrain from adopting them because of some reasons. One of the common reasons claims that the intricacies of these techniques make engines too complex to maintain. Another major barrier is the lack of widely accepted architectures and libraries of these techniques, which leads to the adoption often starting from scratch with lots of effort. In this paper, we propose Intel Big Data Analytic Toolkit (BDTK), an open-source C++ acceleration toolkit library for analytical data processing engines. BDTK provides lightweight, easy-to-connect, reusable components with interoperable interfaces to support query compilation and hardware accelerators. The query compilation in BDTK leverages vectorized execution and data-centric code generation to achieve high performance. BDTK could be integrated into different engines and helps them to adapt query compilation and hardware accelerators to optimize performance bottlenecks with less engineering effort.

An IoT System and MODIS Images Enable Smart Environmental Management for Mekong Delta

The smart environmental management system proposed in this work offers a new approach to environmental monitoring by utilizing data from IoT stations and MODIS satellite imagery. The system is designed to be deployed in vast regions, such as the Mekong Delta, with low building and operating costs, making it a cost-effective solution for environmental monitoring. The system leverages telemetry data collected by IoT stations in combination with MODIS MOD09GA, MOD11A1, and MCD19A2 daily image products to develop computational models that calculate the values land surface temperature (LST), 2.5 and 10 (µm) particulate matter mass concentrations (PM2.5 and PM10) in areas without IoT stations. The MOD09GA product provides land surface spectral reflectance from visible to shortwave infrared wavelengths to determine land cover types. The MOD11A1 product provides thermal infrared emission from the land surface to compute LST. The MCD19A2 product provides aerosol optical depth values to detect the presence of atmospheric aerosols, e.g., PM2.5 and PM10. The collected data, including remote sensing images and telemetry sensor data, are preprocessed to eliminate redundancy and stored in cloud storage services for further processing. This allows for automatic retrieval and computation of the data by the smart data processing engine, which is designed to process various data types including images and videos from cameras and drones. The calculated values are then made available through a graphic user interface (GUI) that can be accessed through both desktop and mobile devices. The GUI provides real-time visualization of the monitoring values, as well as alerts to administrators based on predetermined rules and values of the data. This allows administrators to easily monitor the system, configure the system by setting alerting rules or calibrating the ground stations, and take appropriate action in response to alerts. Experimental results from the implementation of the system in Dong Thap Province in the Mekong Delta show that the linear regression models for PM2.5 and PM10 estimations from MCD19A2 AOD values have correlation coefficients of 0.81 and 0.68, and RMSEs of 4.11 and 5.74 µg/m3, respectively. Computed LST values from MOD09GA and MOD11A1 reflectance and emission data have a correlation coefficient of 0.82 with ground measurements of air temperature. These errors are comparable to other models reported in similar regions in the literature, demonstrating the effectiveness and accuracy of the proposed system.

Plasma glycoproteomic biomarkers identify metastatic melanoma patients with reduced clinical benefit from immune checkpoint inhibitor therapy.

The clinical success of immune-checkpoint inhibitors (ICI) in both resected and metastatic melanoma has confirmed the validity of therapeutic strategies that boost the immune system to counteract cancer. However, half of patients with metastatic disease treated with even the most aggressive regimen do not derive durable clinical benefit. Thus, there is a critical need for predictive biomarkers that can identify individuals who are unlikely to benefit with high accuracy so that these patients may be spared the toxicity of treatment without the likely benefit of response. Ideally, such an assay would have a fast turnaround time and minimal invasiveness. Here, we utilize a novel platform that combines mass spectrometry with an artificial intelligence-based data processing engine to interrogate the blood glycoproteome in melanoma patients before receiving ICI therapy. We identify 143 biomarkers that demonstrate a difference in expression between the patients who died within six months of starting ICI treatment and those who remained progression-free for three years. We then develop a glycoproteomic classifier that predicts benefit of immunotherapy (HR=2.7; p=0.026) and achieves a significant separation of patients in an independent cohort (HR=5.6; p=0.027). To understand how circulating glycoproteins may affect efficacy of treatment, we analyze the differences in glycosylation structure and discover a fucosylation signature in patients with shorter overall survival (OS). We then develop a fucosylation-based model that effectively stratifies patients (HR=3.5; p=0.0066). Together, our data demonstrate the utility of plasma glycoproteomics for biomarker discovery and prediction of ICI benefit in patients with metastatic melanoma and suggest that protein fucosylation may be a determinant of anti-tumor immunity.

Abstract 5314: A peripheral blood-based glycoproteomic predictor of checkpoint inhibitor treatment benefit in advanced non-small cell lung cancer

Abstract Background: Protein glycosylation is the most common and complex form of post-translational protein modification. Glycosylation profoundly affects protein structure, conformation, and function. The elucidation of the potential role of differential protein glycosylation as biomarkers has been limited by the technical complexity of generating and interpreting this information. We have recently established a novel, powerful platform that combines liquid chromatography-mass spectrometry with a proprietary artificial-intelligence-based data processing engine that allows, for the first time, highly scalable interrogation of the glycoproteome. Here we report the performance of this platform to predict likely benefit from immune-checkpoint inhibitor (ICI) therapy in advanced non-small cell lung cancer (NSCLC). Methods: Our platform was utilized to assess 532 glycopeptide (GP) and peptide signatures representing 75 serum proteins in pretreatment blood samples from a cohort of 123 individuals (54 females, 69 males, age range 30 to 88 years). Inclusion criteria were a diagnosis of unresectable stage 3 or 4 NSCLC, treatment with pembrolizumab monotherapy (26 patients), or treatment with combination pembrolizumab-chemotherapy (97 patients). Overall survival (OS) data were available for all patients. Results: An ensemble multivariable-model-based glycoproteomic classifier consisting of 7 GP and non-glycosylated peptide biomarker features selected from a generalized additive model for OS was developed using ≈2/3rds of the full cohort (n=88) and validated in the remainder of patients (n=35). The classifier yielded similar statistical significance in Cox regression analysis for separating patients who are likely to benefit from ICI therapy from those who are not, to accurately predict likely ICI benefit with a sensitivity of >95% while performing at a specificity of 33% to predict those who are unlikely to benefit. Results were further analyzed in patients with either non-squamous or squamous NSCLC with first-line therapy (n=98). The classifier yielded a hazard ratio (HR) for prediction of likely ICI benefit of 3.6 with median OS of 13.9 vs. 4.2 months, and of 3.5 with median OS of 13.5 vs. 4.5 months in the entire cohort and the first-line treated patients, respectively. Conclusions: The glycoproteomic classifier described here predicts with high sensitivity which patients are likely to benefit from ICI therapy. In addition to potentially reducing the use of ICIs in a safe manner in patients who would be unnecessarily subjected to possible adverse drug reactions, our classifier simultaneously has the potential of reducing the burden of health care expenditures. Our results indicate that glycoproteomics holds a strong promise as a predictor for ICI treatment benefit which appears to significantly outperform other currently pursued biomarker approaches. Citation Format: Klaus Lindpaintner, Chad Pickering, Alan Mitchell, Gege Xu, Xin Cong, Daniel Serie. A peripheral blood-based glycoproteomic predictor of checkpoint inhibitor treatment benefit in advanced non-small cell lung cancer. [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 5314.

공장 대기오염 모니터링 응용의 규모 확대를 위한 컨테이너화된 접근법

The deterioration of air quality stems from air pollutants emitted by industrial sectors. Government agencies in many countries regulate the total amount of emissions released during production, necessitating a continuous monitoring system to measure specific air pollutants. However, the existing systems are tailored to a specific plant, making it difficult to incorporate sophisticated components like streaming data processing engines. To address this issue, we propose a containerized approach that bundles rich components for continuous emission monitoring and enables rapid deployments in a new plant. We demonstrate the effectiveness of our approach using datasets obtained from real-world factory production.

MDSuite: comprehensive post-processing tool for particle simulations

Particle-Based (PB) simulations, including Molecular Dynamics (MD), provide access to system observables that are not easily available experimentally. However, in most cases, PB data needs to be processed after a simulation to extract these observables. One of the main challenges in post-processing PB simulations is managing the large amounts of data typically generated without incurring memory or computational capacity limitations. In this work, we introduce the post-processing tool: MDSuite. This software, developed in Python, combines state-of-the-art computing technologies such as TensorFlow, with modern data management tools such as HDF5 and SQL for a fast, scalable, and accurate PB data processing engine. This package, built around the principles of FAIR data, provides a memory safe, parallelized, and GPU accelerated environment for the analysis of particle simulations. The software currently offers 17 calculators for the computation of properties including diffusion coefficients, thermal conductivity, viscosity, radial distribution functions, coordination numbers, and more. Further, the object-oriented framework allows for the rapid implementation of new calculators or file-readers for different simulation software. The Python front-end provides a familiar interface for many users in the scientific community and a mild learning curve for the inexperienced. Future developments will include the introduction of more analysis associated with ab-initio methods, colloidal/macroscopic particle methods, and extension to experimental data.

A Practical Yet Accurate Real-Time Statistical Analysis Library for Hydrologic Time-Series Big Data

Using different statistical analysis methods to examine hydrologic time-series data is the basis of accurate hydrologic status analysis. With the wide application of the Internet of Things and sensor technologies, traditional statistical analysis methods are unable to meet the demand for real-time and accurate hydrologic data analysis. The existing mainstream big-data analysis platforms lack analysis methods oriented to hydrologic data. In this context, a real-time statistical analysis library based on the new generation of big data processing engine Flink, called HydroStreamingLib, was proposed and implemented. Furthermore, in order to prove the efficiency and handiness of the proposed library, a real-time statistical analysis system of hydrologic stream data was developed based on the concepts available in the proposed library. The results showed that HydroStreamingLib provides users with an efficient, real-time statistical verification method, thus extending the application capabilities of Flink Ecology in some specific fields.

Automation of a National Noise Model

The United Kingdom Department for Food and Rural Affairs (Defra) commissioned the design and build of an environmental noise modelling system (NMS). The NMS has been developed to support Defra in developing its environmental noise evidence base by preparing national road and railway noise models. This paper discusses the automation aspects of the NMS with respect to the various elements required for a noise model. This includes the development of sophisticated data processing engines and the techniques utilised to automate the delivery of a 3-dimensional noise propagation environment, and road and railway noise emissions. It highlights several innovative methods which are now possible due to the data landscape in England and how through open data standards the NMS has the potential to improve the fidelity of national noise models.

A resource occupancy ratio-oriented load balancing task scheduling mechanism for Flink

Flink is regarded as a promising distributed data processing engine for unifying bounded data and unbounded data. Unbalanced workloads upon multiple workers/task managers/servers in the Flink bring congestion, which will lead to the quality of service (QoS) decreasing. The balanced load distribution could efficiently improve QoS. Besides, existing works are lagging behind the current Flink version. To distribute workloads upon workers evenly, a resource-oriented load balancing task scheduling (RoLBTS) mechanism for Flink is proposed. The capacities of CPU, memory, and bandwidth are taken into consideration. Based on the barrel principle, the memory, and the bandwidth are respectively selected to model the resource occupancy ratio of the physical node and that of the physical link. On the based of modeled resource occupancy ratio, the data processing of load-balancing resource usage in Flink is formulated as a quadratic programming problem. Based on the self-recursive calling, a RoLBTS algorithm for scheduling task-needed resources is presented. Trough the numerical simulation, the superiority of our work is evaluated in terms of resource score, the number of possible scheduling solutions, and resource usage ratio.

Development of an Artificial Intelligence-Supported Hybrid Data Management Platform for Monitoring Depression and Anxiety Symptoms in the Perinatal Period: Pilot-Scale Study

One of the forces driving science and industry is machine learning, but the proliferation of Big Data necessitates paradigm shifts from conventional approaches in applying machine learning techniques to this massive amount of data with varying velocity. Computers are now capable of accurately diagnosing a variety of medical conditions thanks to the availability of immense healthcare datasets and advancements in machine learning techniques. The study’s primary aim is to identify the most compelling questions on anxiety and depression in pregnant women by extracting features through performance-optimized algorithms. In this way, it is aimed to reach the result in a shorter time with fewer questions. The next goal of this work is to create an instant remote health status prediction system for depression and anxiety in pregnant women based on the Apache Spark Big Data processing engine, which concentrates on using machine learning models on streaming Big Data. In this scalable system, the application receives data from pregnant women to forecast the patient’s health condition. It then applies the machine learning algorithm that produces the best results for this dataset. With the assistance of this big data platform, the time-consuming anxiety and depression detection procedure in a pregnant woman can be replaced with a computer-based technique that works in an instant with a respectable amount of accuracy.

INNOVATIONS IN TECHNOLOGY FOR SOCIAL ENGAGEMENT AMONG OLDER ADULTS AND TO IMPROVE CAREGIVER COMMUNICATION AND TRAINING

Abstract Music-based interventions (MBIs) have been shown to benefit persons with dementia. To develop new MBIs and build the evidence base for motor, cognitive, emotional, and behavioral outcomes, accurate measures of music exposure (e.g., content, dose, and duration) are critically important. However, most commercial music services require internet access and/or subscription fees without providing convenient ways for researchers to access music exposure data. Our research team developed the open-source MUsic to Support Engagement and Resilience (MUSER) Android application with an associated data processing engine. Caregivers can use MUSER to deliver individualized music selections via smartphones/tablets, with music exposure data being generated for research use. We used the Kotlin programming language for application development and Java/Python/Microsoft BI for data processing and analytics instrumentation. The MUSER application was tested for music exposure measurement feasibility with 16 users (ten research team members and six family caregivers of persons with dementia) for two weeks between July and November of 2021. Semi-structured interviews were conducted pre-post application testing to gather qualitative and quantitative data to enrich the understanding of user experiences. Family caregivers had a mean age of 57.8 (SD=26.2; range 20-77). The average number of listening days was 6 (SD=2.5), and the average music listening duration was 203 minutes (SD=123.2). Caregivers played an average of 52 (SD=36.9) songs from 10 (SD=4.2) albums involving 34 (SD=36.3) artists. Given the ease of deployment and integrated data collection and reporting capabilities, this application supports collaborative software development among researchers to advance MBI research.

In-Memory Distributed Mosaicking for Large-Scale Remote Sensing Applications with Geo-Gridded Data Staging on Alluxio

The unprecedented availability of petascale analysis-ready earth observation data has given rise to a remarkable surge in demand for regional to global environmental studies, which exploit tons of data for temporal–spatial analysis at a much larger scale than ever. Imagery mosaicking, which is critical for forming “One Map” with a continuous view for large-scale climate research, has drawn significant concern. However, despite employing distributed data processing engines such as Spark, large-scale data mosaicking still significantly suffers from a staggering number of remote sensing images which could inevitably lead to discouraging performance. The main ill-posed problem of traditional parallel mosaicking algorithms is inherent in the huge computation demand and incredible heavy data I/O burden resulting from intensively shifting tremendous RS data back and forth between limited local memory and bulk external storage throughout the multiple processing stages. To address these issues, we propose an in-memory Spark-enabled distributed data mosaicking at a large scale with geo-gridded data staging accelerated by Alluxio. It organizes enormous “messy” remote sensing datasets into geo-encoded gird groups and indexes them with multi-dimensional space-filling curves geo-encoding assisted by GeoTrellis. All the buckets of geo-grided remote sensing data groups could be loaded directly from Alluxio with data prefetching and expressed as RDDs implemented concurrently as grid tasks of mosaicking on top of the Spark-enabled cluster. It is worth noticing that an in-memory data orchestration is offered to facilitate in-memory big data staging among multiple mosaicking processing stages to eliminate the tremendous data transferring at a great extent while maintaining a better data locality. As a result, benefiting from parallel processing with distributed data prefetching and in-memory data staging, this is a much more effective approach to facilitate large-scale data mosaicking in the context of big data. Experimental results have demonstrated our approach is much more efficient and scalable than the traditional ways of parallel implementing.

The essence of online data processing

Data processing systems are a fundamental component of the modern computing stack. These systems are routinely deployed online: they continuously receive the requests of data processing operations, and continuously return the results to end users or client applications. Online data processing systems have unique features beyond conventional data processing, and the optimizations designed for them are complex, especially when data themselves are structured and dynamic. This paper describes DON Calculus, the first rigorous foundation for online data processing. It captures the essential behavior of both the backend data processing engine and the frontend application, with the focus on two design dimensions essential yet unique to online data processing systems: incremental operation processing (IOP) and temporal locality optimization (TLO). A novel design insight is that the operations continuously applied to the data can be defined as an operation stream flowing through the data structure, and this abstraction unifies diverse designs of IOP and TLO in one calculus. DON Calculus is endowed with a mechanized metatheory centering around a key observable equivalence property: despite the significant non-deterministic executions introduced by IOP and TLO, the observable result of DON Calculus data processing is identical to that of conventional data processing without IOP and TLO. Broadly, DON Calculus is a novel instance in the active pursuit of providing rigorous guarantees to the software system stack. The specification and mechanization of DON Calculus provide a sound base for the designers of future data processing systems to build upon, helping them embrace rigorous semantic engineering without the need of developing from scratch.

Research on Application of Computer Big Data Technology in Real-time Monitoring System of Bridge Structure Safety

The bridge structure safety real-time monitoring system based on big data evaluates the bridge safety status and real-time warning under the comprehensive consideration of various factors that affect the service performance of the bridge. The system platform uses a distributed file system with high fault tolerance and a parallel data processing engine with high computational efficiency. It has high reliability, availability and storage efficiency, and is easy to expand. Furthermore, the multi-factor analysis method is used to fully explore the implicit correlation between the real-time data of the various sensors of the bridge, and the bridge service performance evaluation model is accurately established by analysing the data correlation, and real-time warning of the bridge safety status is given. In addition, the system platform uses a variety of model verification methods to evaluate the effectiveness of the model to ensure the reliability of the big data analysis method.